Enable parallel CvtColorLoop for all platforms

modules/imgproc/src/color.cpp

@@ -157,46 +157,38 @@ template<> struct ColorChannel<float>
 ///////////////////////////// Top-level template function ////////////////////////////////
 
 template <typename Cvt>
-class CvtColorLoop_Invoker : 
+class CvtColorLoop_Invoker : public ParallelLoopBody
-    public ParallelLoopBody
 {
     typedef typename Cvt::channel_type _Tp;
 public:
-    
+
     CvtColorLoop_Invoker(const Mat& _src, Mat& _dst, const Cvt& _cvt) :
         ParallelLoopBody(), src(_src), dst(_dst), cvt(_cvt)
     {
     }
-    
+
     virtual void operator()(const Range& range) const
     {
-        int i = range.start;
+        const uchar* yS = src.ptr<uchar>(range.start);
-        const uchar* yS = src.data + src.step * i;
+        uchar* yD = dst.ptr<uchar>(range.start);
-        uchar* yD = dst.data + dst.step * i;
 
-        for ( ; i < range.end; ++i, yS += src.step, yD += dst.step )
+        for( int i = range.start; i < range.end; ++i, yS += src.step, yD += dst.step )
             cvt((const _Tp*)yS, (_Tp*)yD, src.cols);
     }
-    
+
 private:
-    const Mat src;
+    const Mat& src;
-    Mat dst;
+    Mat& dst;
-    const Cvt cvt;
+    const Cvt& cvt;
-    
+
     CvtColorLoop_Invoker(const CvtColorLoop_Invoker&);
     const CvtColorLoop_Invoker& operator= (const CvtColorLoop_Invoker&);
 };
 
-template <typename Cvt> 
+template <typename Cvt>
 void CvtColorLoop(const Mat& src, Mat& dst, const Cvt& cvt)
 {
-    Range range(0, src.rows);
+    parallel_for_(Range(0, src.rows), CvtColorLoop_Invoker<Cvt>(src, dst, cvt));
-    CvtColorLoop_Invoker<Cvt> invoker(src, dst, cvt);
-#if defined(_MSC_VER) || defined(__APPLE__)
-    parallel_for_(range, invoker);
-#else
-    invoker(range);
-#endif
 }
 
 ////////////////// Various 3/4-channel to 3/4-channel RGB transformations /////////////////
@@ -2172,7 +2164,7 @@ static void Bayer2RGB_VNG_8u( const Mat& srcmat, Mat& dstmat, int code )
     bool haveSSE = cv::checkHardwareSupport(CV_CPU_SSE2);
     #define _mm_absdiff_epu16(a,b) _mm_adds_epu16(_mm_subs_epu16(a, b), _mm_subs_epu16(b, a))
 #endif
-    
+
     for( int y = 2; y < size.height - 4; y++ )
     {
         uchar* dstrow = dst + dststep*y + 6;
@@ -2659,7 +2651,7 @@ static void Bayer2RGB_VNG_8u( const Mat& srcmat, Mat& dstmat, int code )
                 x2 = _mm_merge_epi16(t0, x0);
 
                 uchar R[8], G[8], B[8];
-                
+
                 // Make sure there is no sign bit in the 16 bit values so they can saturate correctly
                 x1 = _mm_and_si128(x1, smask);
                 x2 = _mm_and_si128(x2, smask);
@@ -3180,7 +3172,7 @@ struct RGBA2mRGBA
             _Tp v1 = *src++;
             _Tp v2 = *src++;
             _Tp v3 = *src++;
-            
+
             *dst++ = (v0 * v3 + half_val) / max_val;
             *dst++ = (v1 * v3 + half_val) / max_val;
             *dst++ = (v2 * v3 + half_val) / max_val;
@@ -3205,7 +3197,7 @@ struct mRGBA2RGBA
             _Tp v2 = *src++;
             _Tp v3 = *src++;
             _Tp v3_half = v3 / 2;
-            
+
             *dst++ = (v3==0)? 0 : (v0 * max_val + v3_half) / v3;
             *dst++ = (v3==0)? 0 : (v1 * max_val + v3_half) / v3;
             *dst++ = (v3==0)? 0 : (v2 * max_val + v3_half) / v3;
@@ -3568,7 +3560,7 @@ void cv::cvtColor( InputArray _src, OutputArray _dst, int code, int dcn )
         case CV_BayerBG2BGR: case CV_BayerGB2BGR: case CV_BayerRG2BGR: case CV_BayerGR2BGR:
         case CV_BayerBG2BGR_VNG: case CV_BayerGB2BGR_VNG: case CV_BayerRG2BGR_VNG: case CV_BayerGR2BGR_VNG:
             {
-                if (dcn <= 0) 
+                if (dcn <= 0)
                     dcn = 3;
                 CV_Assert( scn == 1 && dcn == 3 );